Episode 03

The Problem This Episode Solves

 

If market structure is defined discretionarily, it isn’t structure at all.

 

Any structural framework that cannot be applied consistently cannot be measured. And if it cannot be measured, it cannot be conditioned on, compared across datasets, or meaningfully integrated into a trading system. At that point, it becomes narrative — not information.

 

This episode exists to eliminate that problem.

 

By the end of this discussion, market structure should no longer feel like something you “recognise” on a chart, but something you mechanically identify. We’ll move away from subjective labels and toward finite, testable market regimes that can be used to condition probabilities without hindsight bias.

 

This is Episode 03 of the Quant-Inspired Trading Strategy series. We are still in the orientation phase, continuing the feature-definition process introduced in Episode 02.

​

Market Regimes as Market State

 

A market regime is not a forecast, and it is not a signal.

 

It is a state — a recurring condition under which price behaves differently, features interact differently, and outcomes distribute differently relative to other states.

 

Crucially, a regime does not predict what will happen, it conditions what is more or less probable.

 

A useful analogy is weather. The weather does not dictate what you wear, but it meaningfully constrains your choices. In extreme heat, certain outcomes become unlikely. Markets behave the same way. A regime doesn’t guarantee an expansion or a reversal — but it changes the likelihood landscape.

 

This distinction matters because most traders misuse structure as a forecasting tool, rather than a conditioning variable.

 

Why Regimes Are Necessary in Data

​

Without regime differentiation, all observations are implicitly treated as equivalent.

 

This is a problem.

 

A setup might show a 50% win rate across all data, but once you separate that data by regime, you may find that the same setup performs exceptionally well in one state and poorly in another. Without regimes, these contradictions are averaged away, creating the illusion of inconsistency.

 

Regimes don’t make a strategy “work.”

They explain why it works sometimes and fails other times.

​

Why Subjective Structure Fails

 

Most traders rely on descriptive labels such as “trending,” “ranging,” or “choppy.” These labels feel intuitive, but they fail under scrutiny.

 

They are defined differently by different traders, defined differently by the same trader at different times, and often only assigned after the outcome is known. Because of this, they cannot be tested. Any framework that changes its definition depending on context is not a framework — it’s a justification tool.

 

If two traders cannot independently label the same chart in the same way, the structure is not externally observable. And if it cannot be observed consistently, it cannot be measured.

 

Criteria for a Valid Market Regime

​

For a market regime to be useful, it must meet several strict conditions.

 

It must be finite or binary rather than continuous. It must be externally observable so that another trader could apply the same label. It must be stable across time, identifiable in live conditions rather than retrospectively, and behaviorally relevant — meaning it plausibly influences how price interacts with features.

 

Anything that fails these criteria is unsuitable for statistical work.

​

How Structure Is Defined Mechanically

​

Rather than inventing new structural labels, I adopted an existing conceptual framework: market maker models.

 

Market maker models describe whether bullish or bearish expansion is conditionally more probable, based on prior displacement through liquidity. They are not entry models, signals, or guarantees. They describe market state.

 

A model is initiated when price strongly displaces through swing liquidity on a given timeframe. Strong displacement is defined by a candle close that decisively invalidates an opposing confluence, in this case being swing liquidity, rather than merely wicking into it. Weak displacement implies rejection and is thus insufficient.

 

Once initiated, the model remains valid until it is invalidated — either by opposing displacement, by respect of an external opposing confluence, or by initiation of a model in the opposite direction.

 

This definition is mechanical, observable, and testable.

​

Timeframe Relativity and Structural Alignment

 

Market structure is not uniform across timeframes.

 

A lack of structure on one timeframe does not imply a lack of structure on all others. Higher timeframes exert greater influence, but lower timeframes still matter for execution and interaction.

 

Rather than labeling structure independently on each timeframe, I define regimes based on alignment and lack of alignment across timeframes, relative to my entry logic.

 

This produces five finite regimes:

 

• A lack of weekly structure and below

• A lack of 4-hour structure and below

• A lack of hourly structure and below

• Structure in favor of the entry logic

• Structure against the entry logic

​

“In favor” and “against” are not directional labels. They are defined relative to the trade model being executed.

​

Why These Regimes Matter

 

Two identical expansions can occur under different regimes depending on the behavior that preceded them. The outcome may look the same on the right side of the chart, but the probabilistic context is not.

 

Structure is defined by prior behavior, not future results. This is why regimes cannot be assigned retrospectively. If a regime only becomes clear after the move has completed, it has no value in live trading.

 

When regimes are defined mechanically, they allow probabilities to be conditioned honestly. Variance becomes explainable rather than mysterious, and data contradictions disappear.

 

Closing Thoughts

​

Market structure is not a prediction.

 

It is a state.

 

When defined mechanically, market regimes allow you to remove subjectivity, eliminate hindsight bias, and build strategies that are testable rather than narrative-driven.

 

In the next phase of this series, we move out of orientation and into hypothesis formation, where these regimes are no longer theoretical — they are measured.